Autonomous navigation Mecanum wheel robot

I make Autonomous navigation robot using move_base and amcl package with ROS.

Use map created by gmapping(https://youtu.be/9EaXi4bh2s0)

The first video is autonomous navigation before tuning.

The second video is color tracking.(Not using ROS)

The third video is attitude control with IMU.

The fourth video is autonomous navigation after tuning.

The parameter of move_base after tuning has been uploaded to github.

https://github.com/ranbo0311/Mecanum_robot.git

Parameters refer to these codes : https://github.com/zaki0929/raspimouse_navigation_2.git

 

Robot

Size/Weight

 -W x L x H : 200 x 240 x 170 mm

 -1.5 kg

Controller

 -Raspberry Pi 3 B+ (Raspbeian Stretch + ROS kinetic)

 -4 x  Arduino pro mini (MD, Analog sensor + Ultrasonic sensor, 2 x Rotary encoder)

 -PC (Windows 10 + Virtualbox[Ubuntu 16.04 + ROS kinetic])

Hardware

-4 x Mecanum wheel 60 mm

Materials

-Polycarbonate plate (Body)

-ABS (3D printer : https://www.tiertime.com/up-mini-2/)

Actuators

-4 x DC motor (RE-260RA + 74 : 1 gear box)

-2 x servo motor (SG-90)

Sensors/Other devices

-8 x Ultrasonic sensor (HC-SR04)

-1 x Temperature sensor (LM61CIZ)

-1 x IMU (MPU-6050)

-1 x camera (picamera v2)

-1 x Lidar (YDLIDAR X4)

-4 x Rotary encoder (https://www.pololu.com/product/3499)

-4 x Current sensor (ACS712ELCTR-05B-T)

 -1 x Ring LED (Neo Pixel Ring)

 -2 x Motor driver IC (TB6612FNG)

 

Youtube(My channel):https://www.youtube.com/channel/UCgEETDZfbyr0_9rYclqnPsw

Autonomous navigation

  • Actuators / output devices: DC Motors, Servo motors
  • Control method: autonomous
  • CPU: Raspberry Pi, PC, Arduino
  • Programming language: Python, C++
  • Sensors / input devices: YDLIDAR X4, Rotary encoders, IMU sensor, camera

This is a companion discussion topic for the original entry at https://community.robotshop.com/robots/show/autonomous-navigation-mecanum-wheel-robot

More info!

The results are great, but here on LMR we want details! ;o)

More photos, materials list, lessons learned… keep it coming!

Navigation video

Not bad for a start, but for your “parameter adjustment”:Try increasing radius und scaling factor of the inflation_layer in your global costmap so it stays further away from the corner. The navigation tuning guide[1] helped me alot.

Also I believe your odometry needs to be a little more accurate, it shouldn’t cut the corner that often.

 

[1] http://kaiyuzheng.me/documents/navguide.pdf

Thanks for the advice.

I will adjust it as a reference.

I am not good at expressing

I am not good at expressing it in English, but I would like to update as much as possible.

Your English is good enough, the quality of your post is great!

Can you give cartographer and ydlidar code and configuration?

Good stuff, ranbo!

If you haven’t gotten too deep into voice recognition and TTS already, check out Snips, https://docs.snips.ai/

Cool stuff.
Jeff